# 2005 Ford Transmission PRNDL Logic Pulse... need help.

Discussion in 'Electronic Basics' started by Thomas, Apr 1, 2005.

1. ### ThomasGuest

I am trying to activate a relay when the vehicle is in anything other
than Park. The transmission module sends out 5V @ 120hz with varying
pulse widths. In Park, I read 0.71V with the meter and looks to be
apprx 1/7th of the pulse on the scope (square wave). In Reverse, the
pulse is measured at 1.6V of 1/3 of the pulse with higher voltages and
pulse widths in N,D,2. I've tried using this to enable a small dc
relay, but only get chatter (even in Park). I've been told I could do
the following...

RC Integrator - Unity-Gain OpAmp Buffer - Comparator

How would this work? Any help is greatly appreciated.

2. ### Andrew HolmeGuest

The RC integrator will smooth-out the pulses, giving you a DC voltage
proportional to the pulse width.

Your meter is doing the same thing i.e. reading the average level.

The comparator gives you a digital output indicating whether the voltage is
above or below some threshold. You'll need positive feedback (hysteresis)
around the comparator to cope with residual ripple.

There are two tweak-factors to play with:
1. RC time constant
2. Amount of hysteresis

Long time constant = more smoothing = slower response to change
Short time constant = fast response, but more residual ripple on DC voltage
= more hysteresis required

Given the application, a time constant of up to 1 second should be OK, and
would provide effective smoothing at 120Hz.

You won't need a unity gain op-amp buffer if the RC filter output is the
only thing connected to the inverting input of the comparator; and the
threshold preset, and postive feedback resistor, are connected to the
non-inverting input.

3. ### ThomasGuest

Is there any way you could assist in the selection of the correct
components or point me in the right direction as to the right RC
Integrator? Thanks for the input!

4. ### Andrew HolmeGuest

The time constant depends on the product R*C. Many different value pairs
are possible; however, you might start with (say) 100k and 10uF. Look at
the signal on an oscilloscope and experiment with the values. Large values
for R will minimise loading of the input signal. You want the output to be
a steady DC level with minimal ripple. Increasing R*C will reduce ripple.
By minimal ripple, I mean relatively small (say <5%) compared to the level
changes produced by moving the gear selector. This enables the comparator
to distinguish between ripple and gear shifts.

5. ### JamieGuest

you need a Window Comparator. the output of the comparator
can operate a Time Off Delay circuit. each time it triggers
it will recharge the Time off or initially start it.

the whole thing could be built around a simple thing like a
LM324 which has 4 op-amps in one package.
2 of them used as the window comparator.
#3 a slight Time on delay buffer, #4, A time off Delay comparator.

the time off delay is tailured to hold the output steady between